| 摘要: |
| 针对现有组合导航系统易被干扰欺骗以及姿态求解精度不足的问题，设计了惯性测量单元(IMU)与偏振光传感器组成的航姿参考系统(AHRS)。同时，考虑到传统的姿态求解方法精度不高，提出了一种用于仿生导航无人机航姿求解的混合滤波方法。将Mahony滤波后的姿态值作为系统观测量，再结合扩展卡尔曼滤波(EKF)实现传感器数据的深层融合，以获得高精度的姿态角信息。实验结果表明：在静态环境下采用混合滤波方法求解的姿态值能有效滤除偏振光传感器和加速度计内部噪声干扰，其稳定性明显优于两种方法各自求解时的情况；在动态实验中该方法能有效抑制单独采用Mahony滤波时存在的超调问题，表现出更高的动态解算精度，从而为偏振光组合导航系统提供了更精确的姿态估计信息。 |
| 关键词: 组合导航系统 偏振光传感器 混合滤波 无人机 姿态求解 |
| DOI： |
|
| 基金项目:国家重点基础研究发展计划(973计划) (2011CB302101，2011CB302105)；中央高校基本科研业务费专项基金（DUT19LAB11） |
|
| A Hybrid Filter Method for Attitude Determination of UAV in Bionic Navigation |
| JIN Ren-cheng,XIE Lin-da,WEI Wei,WEI Yan-zhao |
| (School of Mechanical Engineering, Dalian University of Technology, Dalian 116024, China) |
| Abstract: |
| Aiming at the problems that not only are the existing integrated navigation systems easy to be disturbed and deceived, but also the accuracy of attitude determination is insufficient, an attitude reference system (AHRS) composed of inertial measurement unit (IMU) and polarized light sensor is designed in this paper. Meanwhile, considering that the accuracy of traditional attitude solving method is not high, a hybrid filter method for attitude determination of UAV in bionic navigation is proposed. The attitude value after Mahony filtering is taken as the system observations, and the sensor data is deeply integrated by combining with extended Kalman filtering (EKF) to obtain the attitude information with high accuracy. The experiment results show that attitude values calculated by the proposed method in the static environment can effectively filter out the internal noise of the polarized-light sensor and the accelerometer, and the stability is significantly better than when the two filter methods are adopted separately. In the dynamic condition, the proposed method can effectively suppress the overshoot problem when Mahony filter is used alone, and show higher dynamic calculation accuracy, and thus can provide more accurate attitude estimation for the polarized light integrated navigation system. |
| Key words: Integrated navigation system Polarized-light sensor Hybrid filter UAV Attitude determination |
